Are windy leaves sunnier? Accounting for within-plot individual-level environmental heterogeneity to prevent biased estimates of between-plot ITV-environmental relationship.

When understanding how intraspecific trait variation (ITV) responds to plot-level environmental gradients at the community scale, studies often overlook within-plot environmental heterogeneities, particularly light variation along the vertical profile of vegetation. This omission can lead to inaccurate estimations of ITV-environmental relationships. Currently, there is no proper analytical framework to address this gap. To address this issue, we propose a conceptual framework that integrates both plot- and within-plot-level environmental factors to analyze ITV-environmental relationships, emphasizing the use of plant individuals as study units and appropriate upscaling to the community scale through a robust sampling design. Furthermore, we applied the proposed framework to empirical data from a subtropical montane cloud forest to reveal the responses of traits from chronic wind stress and light exposure. Our framework can be implemented within a mixed-effects model without substantial drawbacks in terms of statistical power or Type I error rates, as demonstrated by our performance analysis on simulated data. Application to empirical data from a subtropical montane cloud forest further shows that neglecting within-plot light variation may obscure our understanding of plant adaptations to plot-level chronic wind stress. Across response traits, we observed consistent adjustments to light exposure, with more conservative trait expressions in higher light regimes. After accounting for the light effect, both leaf thickness and specific leaf area (SLA) continued to respond to wind stress, but the wind effects were dramatically reduced, especially in SLA. Overall, our results demonstrate that incorporating within-plot environmental variation is critical to accurately quantify ITV–environment relationships and may provide new insights into plant adaptive strategies in trait ecology.